The failure of the gyrocompass occurred downstream of the Yamachiche calling-in point. Before the yaw could be corrected, the vessel had already cleared the Yamachiche anchorage and was heading toward the next anchorage upstream. The pilot had indicated that he intended to anchor in the channel, but the vessel's deep draught forced him to continue on his way. The power failure experienced by the vessel in the Port of Qubec was of short duration and was attributed to a dislocated generator rotor brush. It was possible to put the brush back in place and to restore power. A brief power outage has a negligible effect on the high speed of rotation of a gyrocompass rotor, thereby allowing the compass to remain in azimuth. The worn brushes as well as the replacement brushes used on the generator power unit were not original parts. The information about the first failure was not passed on to the pilot for the Trois-Rivires-Montreal district when he boarded the vessel. The second failure caused a power outage that reduced the speed of the rotor sufficiently to destabilize the gyrocompass. When power was restored to the gyrocompass, it did not stabilize quickly in azimuth. Accordingly, after each course alteration, the gyrocompass was slow to stabilize and exhibited an unknown gyroscopic error. The gyroscopic error, however, decreased with each new course, and the error appeared to be 2E on the Sainte-Anne-de-Sorel course. On each new course, the pilot only had visual observation and the radar to guide him, although the relative-motion radar display did not lend itself to navigation by parallel indexing. As the alignment error of the heading marker was negligible, the lateral radar marks could have been used. The location of buoy S146 to starboard of the heading marker and the vessel's distance of 2.8cable lengths from the south shore were indications that the vessel was running along the north side of the channel. In addition, by reducing the radar gain, the range landmarks on the heading marker could also have been used to manoeuvre in the middle of the channel. On the le du Moine course, however, the range landmarks were astern of the vessel. Several factors affected the conduct of the vessel, including the disruption of the gyrocompass which induced a variable gyroscopic error; the relative-motion radar display which rendered the radar less useful; the absence of buoys S140 and S141; and the reduced visibility which hampered navigation of the vessel by visual observation. In addition, the pilot was considering an overtaking situation between the JEANNIE and the EMERALD STAR. The grounding occurred slightly upstream of the intersection of the Sainte-Anne-de-Sorel and the le du Moine courses, after a turn. Although the pilot was unsure as to the gyroscopic error, he planned a course alteration manoeuvre that involved running along the north side of the channel in readiness for overtaking. Because of the failure of the navigation equipment, the pilot was unable to manoeuvre the vessel with the required precision. After the grounding, the pilot's indication to the EMERALD STAR that the the JEANNIE was on the south side of the channel confirms that the pilot was disoriented following the loss of reference points. Planning for course manoeuvres in the middle of the channel and overtaking in the less confined waters of the Port of Sorel would have reduced the risks of an accident.Analysis The failure of the gyrocompass occurred downstream of the Yamachiche calling-in point. Before the yaw could be corrected, the vessel had already cleared the Yamachiche anchorage and was heading toward the next anchorage upstream. The pilot had indicated that he intended to anchor in the channel, but the vessel's deep draught forced him to continue on his way. The power failure experienced by the vessel in the Port of Qubec was of short duration and was attributed to a dislocated generator rotor brush. It was possible to put the brush back in place and to restore power. A brief power outage has a negligible effect on the high speed of rotation of a gyrocompass rotor, thereby allowing the compass to remain in azimuth. The worn brushes as well as the replacement brushes used on the generator power unit were not original parts. The information about the first failure was not passed on to the pilot for the Trois-Rivires-Montreal district when he boarded the vessel. The second failure caused a power outage that reduced the speed of the rotor sufficiently to destabilize the gyrocompass. When power was restored to the gyrocompass, it did not stabilize quickly in azimuth. Accordingly, after each course alteration, the gyrocompass was slow to stabilize and exhibited an unknown gyroscopic error. The gyroscopic error, however, decreased with each new course, and the error appeared to be 2E on the Sainte-Anne-de-Sorel course. On each new course, the pilot only had visual observation and the radar to guide him, although the relative-motion radar display did not lend itself to navigation by parallel indexing. As the alignment error of the heading marker was negligible, the lateral radar marks could have been used. The location of buoy S146 to starboard of the heading marker and the vessel's distance of 2.8cable lengths from the south shore were indications that the vessel was running along the north side of the channel. In addition, by reducing the radar gain, the range landmarks on the heading marker could also have been used to manoeuvre in the middle of the channel. On the le du Moine course, however, the range landmarks were astern of the vessel. Several factors affected the conduct of the vessel, including the disruption of the gyrocompass which induced a variable gyroscopic error; the relative-motion radar display which rendered the radar less useful; the absence of buoys S140 and S141; and the reduced visibility which hampered navigation of the vessel by visual observation. In addition, the pilot was considering an overtaking situation between the JEANNIE and the EMERALD STAR. The grounding occurred slightly upstream of the intersection of the Sainte-Anne-de-Sorel and the le du Moine courses, after a turn. Although the pilot was unsure as to the gyroscopic error, he planned a course alteration manoeuvre that involved running along the north side of the channel in readiness for overtaking. Because of the failure of the navigation equipment, the pilot was unable to manoeuvre the vessel with the required precision. After the grounding, the pilot's indication to the EMERALD STAR that the the JEANNIE was on the south side of the channel confirms that the pilot was disoriented following the loss of reference points. Planning for course manoeuvres in the middle of the channel and overtaking in the less confined waters of the Port of Sorel would have reduced the risks of an accident. The worn brushes as well as the replacement brushes used on the generator power unit were not original manufacturer's parts. The failure of the generator power unit disrupted the gyrocompass. The pilot was unsure as to the gyroscopic error that the gyrocompass might be exhibiting on the le du Moine course. The location of buoy S146 to starboard of the heading marker and the vessel's distance from the south shore were indications that the vessel was running along the north side of the channel. The le du Moine range marks were not visible because visibility was restricted by sea smoke. The condition of the navigation equipment and the weather conditions did not lend themselves to a precision manoeuvre in a turn on the edge of the channel. The overtaking was not planned outside the confined waters of the channel.Findings The worn brushes as well as the replacement brushes used on the generator power unit were not original manufacturer's parts. The failure of the generator power unit disrupted the gyrocompass. The pilot was unsure as to the gyroscopic error that the gyrocompass might be exhibiting on the le du Moine course. The location of buoy S146 to starboard of the heading marker and the vessel's distance from the south shore were indications that the vessel was running along the north side of the channel. The le du Moine range marks were not visible because visibility was restricted by sea smoke. The condition of the navigation equipment and the weather conditions did not lend themselves to a precision manoeuvre in a turn on the edge of the channel. The overtaking was not planned outside the confined waters of the channel. The failure of the generator power unit which destabilized the gyrocompass produced a variable gyroscopic error that sowed doubt in the pilot's mind. The reduced visibility, the missing buoys and the relative-motion radar display made it impossible to manoeuvre the JEANNIE with the required precision.Causes And Contributing Factors The failure of the generator power unit which destabilized the gyrocompass produced a variable gyroscopic error that sowed doubt in the pilot's mind. The reduced visibility, the missing buoys and the relative-motion radar display made it impossible to manoeuvre the JEANNIE with the required precision.